Skip navigation

Metoclopramide is ineffective in preventing postoperative nausea and vomiting

Metoclopramide has been available for about 40 years, is cheap, and is widely used for treatment and prevention of nausea and vomiting. The implication, therefore, is that we must know all about it. There can be no surprises. Unfortunately for that cosy view of life, a meta-analysis [1] of metoclopramide for preventing postoperative nausea and vomiting shows standard doses to be all but useless.

Search


There was an intensive search of several electronic databases, hand searching, and contact with manufacturers. For inclusion studies had to be randomised comparisons of metoclopramide (any dose or route of administration) with placebo or no treatment, and in adults or children. Trials which treated established nausea and vomiting were not addressed in this review. Data was available in 66 randomised studies in which just over 3000 patients received metoclopramide and another 3000 patients received placebo.

Outcomes


The outcomes were nausea, vomiting or nausea and vomiting. These were examined early (0-6 hours) and late (0-48 hours) after surgery. These outcomes were examined for all trials, and for those where the incidence of nausea and vomiting with placebo fell within certain bands. These were 20-60% for early outcomes and 40-80% for late outcomes. Adults and children were treated separately.

Results


There was a variety of different doses and routes of administration. For many combinations, there was only a single trial. The average incidence of early nausea was 18% with placebo, with a wide range of 3 to 60%. The average incidence of early vomiting with placebo was 31% with a range of 18 to 96%.

The main results, where there were at least three trials or 300 patients, are shown in the Table. In all cases the number needed to treat to prevent one additional case of nausea, vomiting or nausea and/or vomiting was 7 and above for adults and 6 and above for children. There was no evidence of a consistent dose-response.

Table: Main results from meta-analysis of metoclopramide compared with placebo for the prevention of postoperative nausea and vomiting

  Number of Outcome with  
Dose/Route Patients without Time Banding Patients Studies Metoclopramide (%) Placebo (%) NNT (95%CI)

Adults
10 mg iv Nausea Early None 543 10 18 25 no benefit
  Vomiting Early None 538 9 20 31 9.1 (5.5 to 27)
  Nausea and/or vomiting Early None 881 14 32 42 10 (6.2 to 28)
10 mg iv Nausea Early 20-60% 298 5 28 36 no benefit
  Vomiting Early 20-60% 355 7 21 30 11 (5.6 to 140)
10 mg iv Nausea Late None 545 5 48 57 no benefit
  Vomiting Late None 728 8 39 48 10 (6 to 41)
  Nausea and/or vomiting Late None 297 6 49 62 7.3 (4.0 to 41)
10 mg iv Vomiting Late 40-80% 473 5 53 65 7.1 (4.4 to19)
10 mg im Nausea Early None 324 2 8 20 8.8 (5.3 to 26)
  Vomiting Early None 406 3 12 23 9.1 (5.5 to 28)
  Nausea and/or vomiting Early None 388 3 20 34 7.0 (4.4 to 18)

Children
0.25 mg/kg iv Vomiting Early None 510 7 31 48 5.8 (3.9 to 11)
0.25 mg/kg iv Vomiting Early 20-60% 409 5 29 41 7.9 (4.6 to 28)


Adverse effects were extracted from the trials. There was no evidence of a greater incidence of extrapyramidal symptoms, sedation and drowsiness, dizziness and vertigo or headache with metoclopramide at these doses than with placebo.

Comment


This is a fine review. It shows evidence of lack of any clinically significant antiemetic effect of metoclopramide in preventing postoperative nausea and vomiting at standard doses. Given that metoclopramide is widely used for this indication, it suggests that a lot of money is being wasted on an ineffective dose.

That conclusion depends, of course, on the definition of clinical usefulness. This group has previously defined a clinically useful result as a number needed to treat of 5 or below. Metoclopramide could not achieve this benchmark for any combination of dose, outcome in adults or children where there was a worthwhile amount of data. We define this as at least three trials and/or 300 patients studied. Even where point estimates approached 5, the confidence intervals were wide.

Reference:

  1. L Henzi, B Walder, MR Tramèr. Metoclopramide in the prevention of postoperative nausea and vomiting: a quantitative systematic review of randomized placebo-controlled studies. British Journal of Anaesthesia 1999 85: 761-71.
previous or next story in this issue